A patient presents with a rash and fever after recent travel to Bolivia. She was there for one week and returned 3 days ago. She took her antimalarial medication, so could this be dengue fever?

Fever in the returning traveler relies on Bayesian thinking. We start with probabilities based on endemic diseases and then the history and physical examination raises or lowers the probability of each. In this case malaria was prevented with doxycycline, so dengue fever is more likely.

Dengue fever is caused by a virus spread by mosquitos bites. The virus is brought to regional lymph nodes and then spreads throughout the body. It presents commonly with typical viral symptoms of fever, headache, nausea and vomiting, myalgia, maculopapular rash, etc. These findings are not diagnostically specific. Although it is called “breakbone fever,” in fact the finding of myalgia has no value in differentiating dengue from other febrile illnesses.(Low PLoS Neglected Tropical Diseases 2011) In a cohort study, half of patients were misdiagnosed as having upper respiratory infections.(Sirivichayakul, PLoS Neglected Tropical Diseases, 2012) To decide what is specific, consider the pathology.

The pathology at the cellular level is microvascular permeability. The vessels are not necessarily damaged, just leaking.(Nelson Pathology of Emerging Infections, 1998) Plasma leakage presents as edema and petechiae. The more severe the disease (such as dengue hemorrhagic fever or dengue shock syndrome) the more likely these specific features will be prominent. One might see pleural effusions, ascites, and widespread edema. The tourniquet test is a way of identifying microvascular permeability. Inflate a blood pressure cuff to a level between diastolic and systolic blood pressure and leave it there for five minutes. Now the capillaries are stressed. A positive result is the emergence of 10 petechiae per square inch (2.5 cm) on the forearm.(Gregory PLoS Neglected Tropical Diseases 2011) It is around 50% sensitive, with wide variation depending on the study. Its specificity ranges from the high 80s to mid 90s.

Other more specific findings might include hepatomegaly. Hematologic findings such as leukopenia and thrombocytopenia are characteristic of dengue fever and do raise the probability of this disease.

Take home points:

-The differential diagnosis for fever in the returning traveler depends on endemic diseases

-The pathology of dengue fever is vascular permeability, so look for petechiae or edema

-The capillary fragility test is a way of eliciting this vascular permeability


A patient presents with acute testicular pain. You were concerned about spermatic cord (testicular) torsion but find an intact cremasteric reflex. Do you send the patient home?


In 1984 Rabinowitz introduced the cremasteric reflex as diagnostic of spermatic cord torsion. Stroking the inner thigh near the testicle produces ipsilateral contraction of the cremasteric muscle, with retraction of the testis. About half of newborns do not have this reflex, but by the age of 3, it is thought that 100% of boys with normal testicles have this reflex(Caesar RE J Urol. 1994). Elderly men can lose the reflex (DeJong’s The Neurologic Examination)


A hundred years ago this sign was known for being prominent in sciatica, and offered a theoretical means of testing L1 and L2 function, but it was more of a curiosity than a critical part of physical examination. Interestingly, in one of the early case reports of spermatic torsion from 1894, the cremasteric reflex was performed but its significance would not be appreciated for another 90 years (Warbasse JP. Ann Surg. 1894)


For about two decades after Rabinowitz’ 1984 article, the cremasteric reflex was thought to be 100% sensitive for spermatic cord torsion. Van Glabeke’s report in 1999 of only 60% sensitivity should have put that illusion to rest, but the significance of his findings was not always appreciated in subsequent textbooks and lectures. Case reports continued to accumulate illustrating limitations in the sensitivity of the cremasteric reflex for spermatic cord torsion, which should not have surprised anyone. What made Van Glabeke’s report most convincing is that they operated on 98% of children with the acute scrotum (544 out of 556 patients). Thus, this article was least susceptible to incorporation bias. Competing studies that showed high sensitivity for the cremasteric reflex often used different gold standards, such as the documented loss of arterial flow by ultrasound.


So how should we use the cremasteric reflex? A prospective study showed that of 12 patients with spermatic cord torsion, only 8 had a positive cremasteric reflex (Boettcher BJU Int 2013). The same group 2 years prior found that it was only 21% sensitive in 19 torsion cases.(Boettcher Urol 2012) The range can vary but omitting extremes, a rough estimate would say that about 2/3 of cases of torsion will have a pathologic cremasteric reflex and1/3 will be normal.


Back to the bedside. Armed with this information, the normal cremasteric reflex does not call off your concern for torsion and you continue your evaluation.


Take Home Points:

An absent or diminished cremasteric reflex is approximately 60-70% sensitive for spermatic cord torsion.

Studies with different gold standards may exaggerate the sensitivity of the cremasteric reflex.